Risk factors for coronary artery disease (CAD) include age, metabolic and cardiovascular factors (hyperlipidemia, hyperglycemia, hypertension) and also certain living habits (diet, cigarette smoking, and physical activity level). Overweight is known to be associated with these factors. More recently, the pattern of fat distribution is emerging as an important contributory variable. Data obtained on participants in the Baltimore Longitudinal Study of Aging (BLSA) have been analyzed. The Body Mass Index or BMI (weight/height2) was used as an index of overweight and the waist:hip circumferential ratio (WHR) as a measure of fat distribution pattern. Both bivariate and multiple regression techniques were used to assess interrelationships among the variables designated as "independent" (age, BMI, WHR, and physical activity level) as well as to assess their influence on the variables designated as "dependent" (serum cholesterol, triglycerides, glucose tolerance, fasting plasma glucose and systolic and diastolic blood pressure). Habitual physical activity level was obtained from detailed activity history questionnaires completed on each visit since 1965 by most participants in the BLSA. The habitual activity level had very little independent effect on coronary risk factors except for the serum cholesterol level. Thus, although intensive physical activity can be shown under experimental laboratory conditions to alter many coronary risk factors in a favorable direction, in this study of a free-living population the effect of activity is minimal. These surprisingly negative results might be explained if physical activity exerts its effects in this population by acting on body weight, so that it loses its independent significance on multivariate analysis. Another caveat is necessary: population studies which include larger numbers of very highly active individuals may show activity effects not demonstrated in this population. This however remains to be shown.